Text Generation
Transformers
PyTorch
English
llama
Eval Results
text-generation-inference
Inference Endpoints

How can it beat vicuna-13b in reasoning ability?

#8
by xiier - opened

The reasoning ability of chatgpt-3.5 is similar to that of vicuna-13b. Orca-mini uses chatgpt-3.5 as a teacher to learn how to think. How can it beat vicuna-13b in reasoning ability?

图片.png

The whole paper Orca research paper https://arxiv.org/abs/2306.02707 with detailed examples and evaluation is about explaining "How it beat Vicuan-13b in reasoning abilities?" Please give a shot, It's one of the best paper out there.

pankajmathur changed discussion status to closed

Sign up or log in to comment